Aj.Chanchai

Journal of Global Antimicrobial Resistance 10 (2017) 47–53 Contents lists available at ScienceDirectJournal of Global Antimicrobial Resistance journal homepage: www.elsevier.com/locate/jgarThe novel mef(C)–mph(G) macrolide resistance genes are conveyed inthe environment on various vectorsYuta Sugimotoa, Satoru Suzukia,*, Lisa Nonakab, Chanchai Boonlac, Nop Sukpanyathamd,Hsin-Yiu Choue, Jer-Horng Wufa Center for Marine Environmental Studies (CMES), Ehime University, Matsuyama, Japanb Department of Microbiology, Dokkyo Medical University School of Medicine, Mibu, Japanc Department of Biochemistry, Faculty of Medicine, Chulalongkorn University, Bangkok, Thailandd Quality Vet Product, Bangkok, Thailande National Taiwan Ocean University, Keelung, Taiwanf Department of Environmental Engineering, Sustainable Environment Research Laboratories (SERL), National Cheng Kung University, Tainan, TaiwanARTICLE INFO ABSTRACTArticle history: Background: The novel tandem genes mef(C) and mph(G) have been reported in marine bacteria in Japan.Received 28 July 2016 This paper aimed to characterise the extent of environmental distribution of mef(C) and mph(G) as well asReceived in revised form 22 March 2017 their dissemination and persistence in aquatic bacterial communities.Accepted 22 March 2017 Methods: Erythromycin-resistant bacteria were isolated from Japan, Taiwan and Thailand aquacultureAvailable online 6 July 2017 sites. The mef(C)–mph(G) genes were detected by PCR. The size of mobile genetic elements conveying mef (C) and mph(G) was examined by Southern blotting. The conjugation rate was assessed by filter mating.Keywords: Results: The mef(C)–mph(G) tandem genes were distributed in erythromycin-resistant isolates fromMacrolide aquaculture seawater in Japan and northern Taiwan and in animal farm wastewater in Thailand. A total ofResistance 29 bacterial isolates were positive for mef(C)–mph(G). The genes were found on vectors of various sizes.mef Partial sequencing of the traI relaxase gene revealed homology with a pAQU1-like plasmid, an IncA/C-mph type plasmid and an SXT/R391 family integrative conjugative element (SRI) as vectors. Thirteen isolatesAsian waters (45%) were positive for traI(pAQU-IncA/C-SRI), whereas the others were negative. The traI(pAQU-IncA/C- SRI)-positive isolates exhibited a higher transfer frequency (10À4–10À5 transconjugants/donor) than traI (pAQU-IncA/C-SRI)-negative isolates (<10À9). Conclusions: These results suggest that mef(C)–mph(G) are coded on various vectors and are distributed among marine and wastewater bacteria in Asian countries. Vectors with traI(pAQU-IncA/C-SRI) play a role in the spread of mef(C)–mph(G). © 2017 International Society for Chemotherapy of Infection and Cancer. Published by Elsevier Ltd. All rights reserved.1. Introduction phosphorylation [4,5]. The mef and mph genes are generally known to function in efflux and phosphorylation, respectively [5]. Erythromycin (ERY) is a member of the macrolide family ofantibiotics, which are commonly used in aquaculture as well as in Recently, Nonaka et al. found a new cluster of mef(C) and mphhuman and veterinary medicine [1,2]. The antibacterial mecha- (G) genes in Photobacterium sp. [6,7]. The tandem-pair arrange-nism of macrolides involves binding to the 50S subunit of the ment of mef(C) and mph(G) is located on plasmid pAQU1 inbacterial ribosome, thus inhibiting protein synthesis [3]. Three Photobacterium damselae subsp. damselae [6]. Although mef(C) andprimary resistance mechanisms are known: inhibition of ERY mph(G) impart lower ERY resistance by themselves [minimumbinding by methylation of the 50S ribosome binding site; efflux ofERY from the cell; and inactivation of ERY by esterification or inhibitory concentrations (MICs) of 32 mg/mL and 256 mg/mL, * Corresponding author. Fax: +81 89 927 8552. respectively], co-expression of these genes results in higher E-mail address: [email protected] (S. Suzuki). resistance (MIC > 512 mg/mL) [7]. As mef(C) and mph(G) are newly identified genes, their genetic diversity, geographic distribution and vectors have not been examined. Previous studies have reported the distribution of genes conferring resistance to tetracycline and sulfamethoxazole in the aquatic environment [8–11]. To characterise the extent of environmental contaminationhttp://dx.doi.org/10.1016/j.jgar.2017.03.0152213-7165/© 2017 International Society for Chemotherapy of Infection and Cancer. Published by Elsevier Ltd. All rights reserved.

48 Y. Sugimoto et al. / Journal of Global Antimicrobial Resistance 10 (2017) 47–53with mef(C) and mph(G) as well as their dissemination and manufacturer’sprotocol. PCR was performed to detect the mef(C),persistence in aquatic bacterial communities, the distribution and mph(G), traI and 16S rRNA genes. The PCR primers for traI [12] candiversity of these genes and their vectors in Asia were investigated. detect genes from pAQU1-like plasmids, IncA/C-type plasmids andAs a result, it was found that the vectors carrying mef(C)–mph(G) SXT/R391 type-integrative conjugative element (SRI). PCR primersare pAQU1 and IncA/C-type plasmids as well as an SXT/R391-type and conditions for mef(C), mph(G) and mef(C)–mph(G) wereintegrative conjugative element (SRI). according to Nonaka et al. [7]. Bacterial phylogenetic positions were identified by sequencing of 16S rRNA gene between the 341f2. Materials and methods and 907r regions [13,14]. PCR for the traI(pAQU-IncA/C-SRI) was performed for mef(C)- and mph(G)-positive strains. PCR products of2.1. Bacterial isolation 16S rRNA and traI were purified using a QIAquick1 PCR Purification Kit (QIAGEN, Venlo, The Netherlands) and were sequenced using an A total of 239 ERY-resistant (ERYR) bacteria were collected from Applied Biosystems 3730 Â 1 DNA Analyzer (Thermo Fisherthe following sites: seawater and fish intestines from a fish farm in Scientific, Waltham, MA) with a BigDye1 Terminator v.3.1 CycleEhime, Japan; pond seawater from a fish farm and seawater in Sequencing Kit (Thermo Fisher Scientific). Sequences were analysedYilan, Taiwan; seawater, canal water, animal farm wastewater and against those in the DNA Data Bank of Japan (DDBJ) (http://www.wastewater treatment plant water from Bangkok, Thailand; pig ddbj.nig.ac.jp/) using the BLAST program. Phylogenetic trees werefarm wastewater and oyster farm seawater from Tainan, Taiwan; constructed based on DNA sequences using Molecular Evolutionaryand seawater, river water, pig farm wastewater and broiler farm Genetics Analysis 5.2 (MEGA5.2) [15]. Phylogenetic analysis of traIwastewater from Trang, Thailand. Sampling sites, conditions and was also performed. Nucleotide sequences of traI were aligned usingsamples are shown in Fig. 1 and Table 1. CLUSTAL W [16], and a neighbour-joining tree was constructed. Marine broth (MB) (Becton Dickinson, Franklin Lakes, NJ), 2.3. Determination of minimum inhibitory concentrationsLuria–Bertani (LB) broth (Becton Dickinson) and LB + 2% NaCl wereused with the addition of 1.5% agar. All agar plates were also The ERY MIC for mef(C)- and mph(G)-positive strains was determined by Etest (bioMérieux, Lyon, France) on Mueller–Hintonsupplemented with ERY at a final concentration of 16 mg/mL. Plates agar plates (Becton Dickinson); in the case of seawater-derived isolates, plates were supplemented with 2% NaCl. Each culture waswere incubated at the ambient water temperature of each site diluted to a McFarland optical density no. 0.5. Isolates were(Japan and Taiwan samples were incubated at 25 C and Thailand incubated overnight at 25 C for Japan and Taiwan strains or atsamples were incubated at 30 C) for 1 day or 2 days. ERYR bacteria 30 C for Thailand strains.were randomly isolated from the agar plates. The temperature andsalinity of water at each site are shown in Table 1. All strains werepreserved in 15% glycerol at À80 C.2.2. PCR and sequencing 2.4. Pulsed-field gel electrophoresis (PFGE) and Southern hybridisation Total bacterial DNA was extracted using a QuickGene DNA mef(C)–mph(G) was originally reported from plasmid pAQU1.Tissue S Kit (Kurabo, Osaka, Japan) according to the The copy number is ca. 1 per cell (1.04 copies/cell; n = 10), which Fig. 1. Map of sampling sites and samples.

Y. Sugimoto et al. / Journal of Global Antimicrobial Resistance 10 (2017) 47–53 49Table 1 No. of mef(C)–mphCharacters of sampling sites and number of strains harbouring mef(C) and mph(G). (G)-positive isolatesSite/origin Date Water Series name Total CFU/mLb ERYR CFU/mL (n = sample or site number)b 0 temperature 0 ( C)/salinitya 0 1Ehime, Japan April 2013 N/D 1JANF N/D N/D 7Cultured fish (red sea bream intestine) May 2013 N/D 2JANF N/D N/D 7 June 2013 N/D 3JANF N/D N/D 0Seawater July 2013 N/D 4JANF N/D N/D 0 December 2013 N/D 6JANF 9.7 Æ 9.0 Â 105 8.2 Æ 13 Â 104 (n = 5) 1Yilan, Taiwan January 2014 N/D 7JANF 6.3 Â 105 4.4 Â 104 (n = 2) 0Grouper pond seawater April 2013 19/34–36 1JAN N/D N/D 2 May 2013 22/33–36 2JAN N/D N/D 1Seawater June 2013 23–24/34–36 3JAN N/D N/DTainan, Taiwan July 2013 24–25/32–36 4JAN N/D N/D 3Pig farm wastewater January 2014 17/35 7JAN 2.1 Æ 0.4 Â 102 1.4 Æ 1.4 Â 101 (n = 4) 5Oyster farm seawater September 2014 27/32 8JAN N/D N/D 1Bangkok, Thailand 0Chicken farm wastewater November 2013 21/25 3-1' 2.4 Â 104 1.5 Â 103 (n = 1)Pig farm wastewater November 2013 21/20 3-4” 4.8 Â 104 1.2 Â 103 (n = 1) 0Seawater November 2013 21/26 3-1” 2.7 Â 104 1.5 Â 103 (n = 1) 0Canal water November 2013 23/30 3-2' 1.5 Â 102 0.5 Â 101 (n = 1)Wastewater treatment plant 0Trang, Thailand Feb. 2010 20/22 TNP N/D N/D 1Seawater Feb. 2010 21/30–34 TNO N/D N/D 0River water 0Pig farm wastewater September 2014 30/0 TRC 9.5 Â 104 8.4 Â 103 (n = 2) 0Broiler farm wastewater September 2014 34/0 TRP 1.8 Â 105 6.2 Â 103 (n = 2) September 2014 30/22–24 GT 9.0 Æ 4.2 Â 102 7.5 Æ 4.1 Â101 (n = 4) 0 September 2014 31/0 TBC (1.3 Æ 1.1) Â 105 2.6 Æ 2.5 Â 104 (n = 6) 0 September 2014 32/0 WWTP 5.7 Â 104 8.7 Â 103 (n = 2) 0 0 September 2015 29/5 TSE 8.5 Â 102 1.7 Â 101 (n = 2) September 2015 28/0 TSR 3.0 Â 103 1.1 Â103 (n = 2) September 2015 18/0 TSPF 5.8 Â 105 7.9 Â 104 (n = 2) September 2015 17–21/0 TSBF 1.5 Â 104 5.2 Â 103 (n = 1)ERYR, erythromycin-resistant; N/D, not determined. a Water temperature and salinity are shown for each water sample. Indication of xx–xx shows range of two to four sites. Salinity unit is psu. b The standard deviation was calculated in the case of n ! 3.was calculated from data in Bien et al. [17]. It is difficult to purify incubated at 25 C for 20 h to allow for conjugation. To select thethe plasmid, thus Southern blotting of total DNA extracted from transconjugants, the filter was suspended in phosphate-bufferedbacterial cells was performed to detect mef(C)–mph(G) and traI. saline (PBS), which was spread on an LB plate containing 100 mg/ Total bacterial DNA was separated into chromosomal DNA mL ERY and 25 mg/mL kanamycin after appropriate dilution. Theand plasmid DNA by PFGE. Following electrophoresis, themigration sizes of mef(C), mph(G) and traI were determined by plate was incubated at 42 C for 2 days. Donors and recipients wereSouthern hybridisation with probes prepared from each PCR unable to grow under these conditions, at which only trans-product. PCR products of mef(C), mph(G) and traI(pAQU-IncA/C- conjugants can grow. The number of donor plus transconjugantSRI) were labelled with digoxigenin using a PCR DIG ProbeSynthesis Kit (Roche Diagnostics, Basel, Switzerland). PFGE and CFUs was determined on LB agar plates with 100 mg/mL ERY and 2%Southern hybridisation were performed according to themethods described by Nonaka et al. [6]. The hybridisation NaCl following incubation at 25 C for 2 days. Although the MIC oftemperature used in this study (Topt = 41 C for all three genes)enabled the detection of 80% of the homologous fragments the recipient strain (E. coli JW0452) is 4 mg/mL, this strain couldaccording to the manufacturer’s protocol (Roche Diagnostics). grow at 64 mg/mL when the cell density is high. Therefore, the ERYProMega-Markers1 Lambda Ladders (Promega Corp., Fitchburg, concentration for screening was 100 mg/mL. After filter mating,WI) were used as size markers. PCR was used to confirm that transconjugants harboured mef(C)–2.5. Filter mating mph(G). The transfer rate was calculated as the number of transconjugant CFUs per donor. Experiments were performed in mef(C)- and mph(G)-positive bacteria were employed as donors triplicate.and Escherichia coli JW0452 Keio Collection (Coli Genetic StockCenter; http://cgsc.biology.yale.edu/KeioList.php) was employed 2.6. Statistical analysisas the recipient. Strain JW0452 was confirmed as ERY-susceptibleand mef(C)- and mph(G)-negative. Filter mating was performed To determine the experimental error range of vector size,according to Nonaka et al. [12]. The donor was cultured with ERY pAQU1 was used as a standard plasmid. In the PFGE-Southernon a filer. The filter was washed with 5 mL of LB medium to remove blotting, pAQU1 (204 052 bp) was detected at a size ofERY from the filter and was placed on an MB plate, which was 232.8 Æ 12.1 kbp in five replicate experiments. Clustering of the migration sizes of mef(C), mph(G) and traI(pAQU-IncA/C-SRI) was performed using the furthest neighbour method [18]. The standard deviations of transfer frequencies of each strain were calculated from triplicate experiments. Differences between high and low transfer frequencies were determined by t-test.

50 Y. Sugimoto et al. / Journal of Global Antimicrobial Resistance 10 (2017) 47–53Table 2List of strains harbouring mef(C) and mph(G).Site/origin Date Strain ID Accession no. Genus ERY MIC (mg/mL) mef(C) mph(G) mef(C)–mph(G)Ehime, JapanCultured fish (red July 2013 4JANF3-O-5 LC191812 Photobacterium sp. >256 + + + December 2013 6JANF1-E-1 LC191813 Vibrio sp. >256 + + + sea bream intestine) December 2013 6JANF2-E-1 LC191814 Vibrio sp. 96 + + + December 2013 6JANF2-E-3 LC191815 Vibrio sp. 16 – + –Seawater December 2013 6JANF4-E-1 LC191816 Photobacterium sp. >256 + + +Yilan, Taiwan December 2013 6JANF4-E-3 LC191817 Photobacterium sp. >256 + + +Grouper pond seawater December 2013 6JANF4-E-4 LC191818 Shewanella sp. >256 + + + December 2013 6JANF5-E-2 LC191819 Vibrio sp. >256 + + +Bangkok, Thailand December 2013 6JANF5-E-3 LC191820 Vibrio sp. >256 + + +Pig farm wastewater January 2014 7JANF1-E-2 LC191821 Vibrio sp. 64 + + + January 2014 7JANF1-E-3 LC191822 Vibrio sp. 48 + + + January 2014 7JANF3-E-1 LC191823 Vibrio sp. >256 + + + January 2014 7JANF3-E-2 LC191824 Vibrio sp. >256 + + + January 2014 7JANF3-E-3 LC191825 Vibrio sp. 24 – + – January 2014 7JANF4-E-2 LC191826 Vibrio sp. >256 + + + January 2014 7JANF5-E-1 LC191827 Vibrio sp. 128 + + + January 2014 7JANF5-E-3 LC191828 Citrobacter sp. >256 + + + June 2013 3JAN1-O-3 LC191829 Pseudoalteromonas sp. >256 + + + January 2014 7JAN2-E-2 LC191830 Vibrio sp. >256 + + + January 2014 7JAN2-E-4 LC191831 Vibrio sp. >256 + + + September 2014 8JAN6-E-1 LC191832 Vibrio sp. >256 + + + November 2013 3-1'-E3 LC191833 Photobacterium sp. >256 + + + November 2013 3-1'-E4 LC191834 Photobacterium sp. >256 + + + November 2013 3-1'-E6 LC191835 Photobacterium sp. >256 + + + November 2013 3-4”-E1 LC191836 Photobacterium sp. 96 + + + November 2013 3-4”-E2 LC191837 Photobacterium sp. >256 + + + November 2013 3-4”-E4 LC191838 Vibrio sp. 64 + + + November 2013 3-4”-E6 LC191839 Photobacterium sp. >256 + + + November 2013 3-4”-E7 LC191840 Photobacterium sp. >256 + + + November 2013 3-1”-E4 LC191841 Photobacterium sp. >256 + + + September 2014 TRP1-in-E-2 LC191842 Proteus sp. >256 + + +ERY, erythromycin; MIC, minimum inhibitory concentration.3. Results and discussion Table 3 traI type in mef(C)- and/or mph(G)-positive isolates and transfer frequency of mef(C)3.1. Detection of mef(C) and mph(G) and mph(G) by filter mating. Among the sampling sites, the occurrence of ERYR bacteria in traI type Strain ID Site Transfer frequencyriver and wastewater was 3.5–36% of the total colony-forming pAQU1bacteria, whereas the occurrence in seawater samples in all pAQU1-like 04Ya311 Kagawa 1.7 Æ 2.8 Â 10À5countries was 1.9–8.4% (Table 1). These showed a significant 6JANF2-E-1 Ehime 6.1 Æ 3.5 Â 10À6difference (P < 0.05), suggesting that ERYR bacteria distribute SXT/R391 family ICE 6JANF2-E-3a Ehime 2.2 Æ 0.0068 Â 10À4widely and the occurrence rate is higher in wastewater compared 6JANF4-E-1 Ehime 1.1 Æ 0.54 Â 10À5with seawater even in aquaculture areas. IncA/C 6JANF4-E-3 Ehime 2.6 Æ 3.5 Â 10À6 Not detected 7JANF3-E-1 Ehime 5.2 Æ 1.1 Â10À9 Of the 239 ERYR isolates examined, PCR indicated that 29 7JANF3-E-2 Ehime 9.6 Æ 8.5 Â 10À9(12.1%) were positive for mef(C), mph(G) and mef(C)–mph(G), 7JANF3-E-3a Ehime <1.5 Â 10À9whilst 2 (0.8%) were positive for mph(G) only. The 31 isolates are 3-4”-E4 Yilan 6.1 Æ 4.4 Â 10À5listed in Table 2 with their strain ID. These 31 isolates included 6JANF4-E-4 Ehime 7.8 Æ 2.7 Â 10À521 isolates from Ehime, 9 from Yilan and 1 from Bangkok, 6JANF5-E-2 Ehime 3.4 Æ 0.82 Â 10À4suggesting that these genes are present in bacteria not only along 6JANF5-E-3 Ehime 3.6 Æ 1.7 Â 10À5the coast of Japan [6,7] but also in waters around Taiwan and 7JANF4-E-2 Ehime 3.0 Æ 0.25 Â 10À6Thailand. Positive isolates were obtained from fish intestine and 7JAN2-E-2 Ehime <3.1 Â10À10seawater in different years, suggesting that mef(C) and mph(G) are TRP1-in-E-2 Bangkok <1.2 Â 10À12persistent in the environment. 3JAN1-O-3 Ehime <1.0 Â 10À11 4JANF3-O-5 Ehime <3.2 Â 10À9 The genera of mef(C)- and mph(G)-positive isolates were widely 6JANF1-E-1 Ehime <3.9 Â 10À10distributed and included Photobacterium (11 isolates), Vibrio 7JANF1-E-2 Ehime <1.1 Â10À10(16 isolates), Shewanella (1 isolate), Pseudoalteromonas (1 isolate), 7JANF1-E-3 Ehime <1.9 Â 10À10Citrobacter (1 isolate) and Proteus (1 isolate) (Table 2). Genera 7JANF5-E-1 Ehime <1.6 Â 10À11harbouring mef(C) and mph(G) included not only marine bacteria 7JANF5-E-3 Ehime <2.6 Â 10À11(Vibrio,Photobacterium, Shewanella and Pseudoalteromonas) but 7JAN2-E-4 Ehime <3.0 Â 10À8also enteric bacteria (Citrobacter and Proteus). The Citrobacter sp. 8JAN6-E-1 Ehime 9.3 Æ 3.6 Â 10À7was isolated from farmed fish in Ehime, Japan, and the Proteus sp. 3-1'-E3 Yilan <1.3 Â 10À10was isolated from pig farm wastewater in Bangkok, Thailand, 3-1'-E4 Yilan <1.9 Â 10À10suggesting that these bacteria were excreted by animals and then 3-1'-E6 Yilan <1.6 Â 10À10entered the wastewater and sea. 3-4”-E1 Yilan <7.2 Â 10À9 3-4”-E2 Yilan <2.1 Â10À10 3-4”-E6 Yilan <3.0 Â 10À10 3-4”-E7 Yilan <1.6 Â 10À10 3-1”-E4 Yilan <2.6 Â 10À10 a mef(C)-negative.

Y. Sugimoto et al. / Journal of Global Antimicrobial Resistance 10 (2017) 47–53 51Fig. 2. (A) Determination of the migration sizes of mef(C), mph(G) and traI(pAQU-IncA/C-SRI) by pulsed-field gel electrophoresis (PFGE) and Southern hybridisation. Four ofthe isolates listed in Table 1 (6JANF1-E-1, 7JANF5-E-3, 7JAN2-E-2 and 7JAN2-E-4) did not yield a PFGE band owing to degradation of the DNA, possibly by endogenous DNaseactivity. (B) Clustering of isolates in (A) by the furthest neighbour method. Of the 29 isolates harbouring mef(C)–mph(G), 23 exhibited a on Southern hybridisation, although it was PCR-positive for mef(C), mph(G) and traI(pAQU-IncA/C-SRI). We hypothesise that mef(C)higher MIC (>256 mg/mL) than isolates possessing only one of and mph(G) could be carried by a low-copy-number vector that isthe genes (16 mg/mL or 24 mg/mL), confirming previous findings not detectable by Southern hybridisation.[7]. When traI(pAQU-IncA/C-SRI) was also detected on the same PFGE gel, the size was the same as that of mef(C) and mph(G). This3.2. Vectors for mef(C) and mph(G) result suggests that the mef(C)–mph(G) tandem is encoded on the same vector that encodes traI(pAQU-IncA/C-SRI). The mef(C) and The mef(C) and mph(G) genes were originally identified on the mph(G) genes are homologous to mef(A) from Tn1207.1 [19] andpAQU1 plasmid, which has a similar transfer mechanism to IncA/ mph(A) from pEK499 [20], respectively. Both of these genes areC-type plasmids [6]. For mass screening of vector(s) encoding mef flanked by other genes that also contribute to macrolide resistance.(C) and mph(G), the migration size of mef(C) and mph(G) was Szczepanowski et al. reported the combination of mph(E) and mrxdetermined using PFGE Southern blotting (Fig. 2). Both of the (E) genes coding for a phosphotransferase and a transmembranegenes in nine Ehime isolates were found to be similar in size to transport protein that together confer macrolide resistance [21].pAQU1, with one exception: 4JANF3-O-5 (400 kbp). The original Different combinations of tandem genes conferring macrolideplasmid pAQU1 can be detected at 232.8 Æ 12.1 kbp (n = 5) in resistance are present in many vectors carried in a broad range ofSouthern blotting. Seven Ehime isolates showed positive bands bacterial species.>700 kbp, suggesting the genes were encoded on the chromo-some. Eight of nine Yilan isolates showed bands of 369–455 kbp, The phylogenetic relationship among traI is shown in Fig. 3. Theand the band from one strain was detected at a size of homology of traI(pAQU-IncA/C-SRI) to previously reported genes306 Æ 10 kbp. This suggests that although the two genes are was as follows: eight isolates (seven from Ehime and one fromconveyed on the same vector at each sampling site in Ehime and Yilan) exhibited 98–100% homology to pAQU1 traI; four EhimeYilan, the vectors differ in each country. Isolates 6JANF2-E-3 and isolates were 99% homologous to ICEVchMex1 [22]; and one isolate7JANF3-E-3 were positive in Southern blotting for all genes, from Ehime was 96% homologous to ICEVflInd1 [23]. Bangkokdespite PCR results indicating that the isolates were mef(C)- isolates exhibited 99% homology to pR148 traI found in IncA/C typenegative (Table 2). This could be attributed to different hybrid- (NC_019380) [24] and pNDM10469 (NC_019158). These resultsisation properties between Southern blotting and PCR. In contrast, suggest that several mobile genetic elements play a role as carrierthe Bangkok isolate (TRP1-in-E-2) did not exhibit a positive band of mef(C)–mph(G) in the environment.

52 Y. Sugimoto et al. / Journal of Global Antimicrobial Resistance 10 (2017) 47–53Fig. 3. Phylogenetic analysis of traI. A neighbour-joining tree was constructed from 522-bp DNA fragments of the traI gene. The number of each node denotes the number oftimes that the tree configuration occurred in 1000 bootstrap trials. The scale bar indicates 0.1 fixed nucleotide substitutions per sequence position. The accession numbers ofsequences used in this analysis are shown in parentheses. ICE, integrative conjugative element.3.3. Transfer frequency of mef(C)–mph(G) and mph(G) are distributed among bacterial communities by a variety of transfer vectors. Transfer of mef(C)–mph(G) was examined by filter matinganalysis (Table 3). Six strains (6JANF2-E-3, 6JANF4-E-1, 3-400-E4, Funding6JANF4-E-4, 6JANF5-E-2 and 6JANF5-E-3) exhibited high transferfrequencies (10À4–10À5 level) similar to that of the original strain, This work was supported in part by grants from KAKENHI,04Ya311 (AB571865) [6], and six other strains exhibited a lower MEXT, Japan [25257402, 16H01782 and 15K07531].transfer frequency (10À6–10À9 level). The high- and low-frequencygroups showed a significant difference (P < 0.05). These results Competing interestssuggest that the transfer of mef(C)–mph(G) is related to the traI(pAQU-IncA/C-SRI) on the vector carrying them. Three traI-positive None declared.strains could not transfer the resistance genes, which might be dueto the lack of other genes essential for conjugal transfer. Ethical approval Almost all of the isolates without traI(pAQU-IncA/C-SRI) were Not required.negative for transfer. One exception (8JAN6-E-1) exhibited atransfer frequency of 9.3 Â 10À7, indicating that this isolate uses a Acknowledgmentsgene transfer system involving an unknown traI or a differentmechanism. With the exception of one isolate (3-400-E4), the Yilan The authors thank Drs M. Usui, H. Takada, Y. Tamura, T.isolates did not transfer mef(C) and mph(G). As shown in Fig. 2, mef Yokokawa and K.-H. Lin for their help with sampling.(C) and mph(G) in the Yilan isolates exhibited a larger migrationsize than pAQU1 and Ehime isolates, suggesting that Yilan isolates Referencescould have a completely different traI allele or use a differentmechanism of transfer. The present work is the first to report the [1] Achard A, Guerin-Faublee V, Pichereau V, Villers C, Leclercq R. Emergence oftransfer of macrolide resistance genes via not only plasmids but macrolide resistance gene mph(B) in Streptococcus uberis and cooperativealso SRIs. effects with rdmC-like gene. Antimicrob Agents Chemother 2008;52:2767–70. In conclusion, the macrolide resistance genes mef(C) and mph [2] Creeper JH, Buller NB. An outbreak of Streptococcus iniae in barramundi (Lates(G) were detected in bacteria isolated from aquaculture seawater calcarifera) in freshwater cage culture. Aust Vet J 2006;84:408–11.in Japan and Taiwan and from pig farm wastewater in Thailand. Themef(C) and mph(G) genes were detected with traI(pAQU-IncA/C- [3] Kohanski MA, Dwyer DJ, Collins JJ. How antibiotics kill bacteria: from targets toSRI) on vectors of various sizes. These results indicate that mef(C) networks. Nat Rev Microbiol 2010;8:423–35.

Y. Sugimoto et al. / Journal of Global Antimicrobial Resistance 10 (2017) 47–53 53 [4] Roberts MC, Sutclifee J, Curvalin P, Jensen LB, Rood J, Seppala H. Nomenclature [15] Tamura K, Peterson D, Peterson N, Stecher G, Nei M, Kumar S. MEGA5: for macrolide and macrolide–lincosamide–streptogramin B resistance deter- molecular evolutionary genetics analysis using maximum likelihood, evolu- minants. Antimicrob Agents Chemother 1999;43:2823–30. tionary distance, and maximum parsimony methods. Mol Biol Evol 2011;28:2731–9. [5] Roberts MC. Update on macrolide–lincosamide–streptogramin, ketolide, and oxazolidinone resistance genes. FEMS Microbiol Lett 2008;282:147–59. [16] Higgins D, Thompson J, Gibson T, Thompson JD, Higgins DG, Gibson TJ. CLUSTAL W. improving the sensitivity of progressive multiple sequence [6] Nonaka L, Maruyama F, Miyakoshi M, Kurakawa K, Masuda M. Novel alignment through sequence weighting, position-specific gap penalties and conjugative transferable multiple drug resistance plasmid pAQU1 from weight matrix choice. Nucleic Acids Res 1994;22:4673–80. Photobacterium damselae subsp. damselae isolated from marine aquaculture environment. Microbes Environ 2012;27:263–72. [17] Bien TLT, Sato-Takabe Y, Ogo M, Usui M, Suzuki S. Persistence of multi-drug resistance plasmids in sterile water under very low concentrations of [7] Nonaka L, Maruyama F, Suzuki S, Masuda M. Novel macrolide-resistance tetracycline. Microbes Environ 2015;30:339–43. genes, mef(C) and mph(G), carried by plasmids from Vibrio and Photobacterium isolated from sediment and seawater of a coastal aquaculture site. Lett Appl [18] Schloss PD, Handelsman J. Introducing DOTUR, a computer program for Microbiol 2015;61:1–6. defining operational taxonomic units and estimating species richness. Appl Environ Microbiol 2005;71:1501–6. [8] Akinbowale OL, Peng H, Barton MD. Antimicrobial resistance in bacteria isolated from aquaculture sources in Australia. J Appl Microbiol [19] Santagati M, Iannelli F, Oggioni MR, Stefani S, Pozzi G. Characterization of a 2006;100:1103–13. genetic element carrying the macrolide efflux gene mef(A) in Streptococcus pneumoniae. Antimicrob Agents Chemother 2000;44:2585–7. [9] Suzuki S, Hoa PTP. Distribution of quinolones, sulfonamides, tetracyclines in aquatic environment and antibiotic resistance in Indochina. Front Microbiol [20] Woodford N, Carattoli A, Karisik E, Underwood A, Ellington MJ, Livermore DM. 2012;3:67. Complete nucleotide sequences of plasmids pEK204, pEK499, and pEK516, encoding CTX-M enzymes in three major Escherichia coli lineages from the[10] Cesare AD, Luna GM, Vignaroli C, Pasquaroli S, Tota S, Paroncini P, Biavasco F. United Kingdom, all belonging to the international O25:H4-ST131 clone. Aquaculture can promote the presence and spread of antibiotic-resistant Antimicrob Agents Chemother 2009;53:4472–82. enterococci in marine sediments. PLoS One 2012;8:e62838. [21] Szczepanowski R, Krahn I, Bohn N, Pühler A, Schlüter A. Novel macrolide[11] Suzuki S, Ogo M, Miller TW, Shimizu A, Takada H, Siringan MAT. Who possesses drug resistance genes in the aquatic environment?: sulfamethoxazole (SMX) resistance module carried by the IncP-1b resistance plasmid pRSB111, isolated resistance genes among the bacterial community in water environment of Metro-Manila, Philippines. Front Microbiol 2013;4:102. from a wastewater treatment plant. Antimicrob Agents Chemother 2007;51:673–8.[12] Nonaka L, Maruyama F, Onishi Y, Kobayashi T, Ogura Y, Hayashi T, et al. Various [22] Burrus V, Quezada-Calvillo R, Marrero J, Waldor MK. SXT-related integrating pAQU plasmids possibly contribute to disseminate tetracycline resistance conjugative element in new world Vibrio cholerae. Appl Environ Microbiol gene tet(M) among marine bacterial community. Front Microbiol 2014;5:152. 2006;72:3054–7. [23] Ahmed AM, Shinoda S, Shimamoto T. A variant type of Vibrio cholerae SXT[13] Muyzer G, Waal ECD, Uitterlinden AG. Profiling of complex microbial element in a multidrug-resistant strain of Vibrio fluvialis. FEMS Microbiol Lett populations by denaturing gradient gel electrophoresis analysis of polymerase 2005;242:241–7. chain reaction-amplified genes coding for 16S rRNA. Appl Environ Microbiol [24] Castillo CSD, Hikima J, Jang H-B, Nho S-W, Jung T-S, Wongtavatchai J, et al. 1993;59:695–700. Comparative sequence analysis of a multidrug-resistant plasmid from Aeromonas hydrophila. Antimicrob Agents Chemother 2013;57:120–9.[14] Lane DJ, Pace B, Olsen GJ, Stahl DA, Sogin ML, Pace NR. Rapid determination of 16S ribosomal RNA sequences for phylogenetic analyses. Proc Natl Acad Sci U S A 1985;82:6955–9.